Disk Drive Partitioning for Automated Storage Tiering
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing storage systems face challenges in efficiently managing data across multiple tiers of storage when not all types of storage devices are available, particularly when Fibre Channel drives are absent, as they struggle to provide the desired performance characteristics and cost efficiency.
Innovation Solution
The system subdivides each disk drive platter into multiple portions with different average response times, allowing frequently accessed data to be stored in faster, more expensive sections and less frequently accessed data in slower, less expensive sections, using software to manage and optimize data placement across similar performance disk drives.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If multiple types of storage devices (Flash, Fibre Channel, SATA) are used to create multiple storage tiers, then data response time performance and cost efficiency are improved, but device complexity and availability requirements worsen
Solution Approach 1:
The patent subdivides a single type of disk drive into multiple virtual storage tiers by partitioning the disk surface into different zones (inner, middle, outer regions). Each zone is assigned different performance characteristics and access priorities through software management, creating multiple storage tiers from a single physical device type. This eliminates the need for multiple different storage device types while maintaining tiered storage functionality.
Solution Approach 2:
The patent makes a single disk drive type perform multiple functions by assigning different zones of the same physical drive to different virtual storage tiers. The same SATA or Fibre Channel drive can simultaneously provide high-performance, medium-performance, and low-performance storage capabilities through software-based zone management, replacing the need for multiple specialized drive types.
2Speed
If data is stored on faster storage tiers (Flash, Fibre Channel), then data response time performance is improved, but storage cost increases
Solution Approach 1:
The patent divides the storage capacity of each disk drive into multiple performance tiers based on physical zone characteristics. Frequently accessed data is automatically placed in higher-performance zones (inner regions with faster access), while less frequently accessed data is placed in lower-performance zones (outer regions). This enables cost-effective tiered storage using a single drive type, avoiding the need to purchase expensive Flash or Fibre Channel drives for all data.
Solution Approach 2:
The patent changes the performance parameters of different zones within the same disk drive by adjusting access patterns, cache allocation, and I/O scheduling priorities. Inner zones are configured with higher performance parameters (faster access times, higher priority caching), while outer zones use lower performance parameters, creating virtual storage tiers with different performance characteristics from identical hardware.
3Device complexity
If a single type of storage device is used, then device complexity and cost are reduced, but the ability to provide multiple storage tiers with different performance characteristics deteriorates
Solution Approach 1:
The patent segments the addressable space of a single disk drive into multiple virtual tiers by partitioning the disk surface into different zones. Each zone is managed independently with different performance characteristics through software control, enabling multiple storage tiers to be created from a single homogeneous storage device type without requiring multiple different drive technologies.
Solution Approach 2:
The patent introduces a software layer (storage manager/director) that acts as an intermediary between the host system and the physical disk drives. This software layer virtualizes the single drive type into multiple performance tiers by managing data placement, access patterns, and zone assignments, thereby providing tiered storage functionality without requiring multiple different hardware drive types.
Data Source
AI summary
Delivering different data response time performance from a plurality of disk drives having similar performance characteristics includes subdividing each disk drive platter of the disk drives into at least two separate portions, where a first portion has a first average response time and the second portion has a second average response time that is greater than the first average response time and includes placing data that is relatively frequently accessed in the first portion of the disk platters of the disk drives to provide a subset of data having a relatively higher data response time performance than other data. Data having a relatively lower data response time performance may be placed on disk drives containing data having a relatively higher data response time performance.